ai engineer · montreal
Hola, I'm Gabri Cebria.
I'm an AI engineer based in Montreal. I build production agent systems and the infrastructure that keeps them running.
I started as a systems engineer at Inmarsat in London on satellite systems, then moved to GTD in Barcelona to work on ESA ground software.
Since then, through my consulting work (Santander, startups and now AI consulting) and my own products, the stack keeps changing but the work rhymes!
I have built my own products along the way too, including Chesskool, a chess training platform I built in 2022 and sold in 2025, and now ClassGenie, an AI language learning platform in pilot with three Spanish schools.
My most recent open source project is Snoopstr, a multi-agent competitive intel tool that runs real cloud browsers in parallel.
Some projects I loved
Snoopstr
Open-source multi-agent competitive intelligence.
You describe your business and four agents run in parallel across real cloud browsers. They pull pricing tiers, landing page persuasion, active Facebook ads, and Instagram signals. Streamed back live to the UI.
ClassGenie
Language learning that runs like a private tutor for every kid.
Average student speaks two minutes in a typical class. ClassGenie turns that into twenty. Agents work in parallel to plan lessons, grade exercises, hold live audio conversations, and adapt to each student. Now used in three Spanish schools.
Chesskool (acquired)
Online chess training platform, sold May 2025.
I built Chesskool in 2022. It grew to over a thousand paying users across LATAM and Europe and hundreds of monthly subscribers. I sold the project in May 2025.
a bit more about me
1991 Cáceres
Born in Cáceres, Spain. Raised as a future tennis superstar (obsession #1).
2003 tennis courts
Didn't make it as a pro but picked up the value of hard work along the way.
2008 chess
Found chess somewhere in there and got pretty into it (obsession #2).
2009-14 Sevilla
Telecommunications engineering degree in Sevilla, BSc + Master's.
2015-2017 Inmarsat, London
Moved to London at 23 for Inmarsat. First year on the RAM team for Global Xpress (GX), the Ka-band high-throughput satellite network, and the I-4 geostationary constellation. Second year I was selected for Inmarsat's graduate program as a systems engineer and researched beam hopping techniques for future GX satellites, optimizing bandwidth allocation around the world.
2018-2020 GTD, Barcelona
After Inmarsat I moved to Barcelona and joined GTD as a software engineer. Built internal web apps for the Ariane rocket launcher controllers operated out of French Guiana, mostly Python on the backend with Vue.js on the front. Also worked on Odyssea, the EU Horizon 2020 Mediterranean marine data platform, building GeoDjango and PostGIS geospatial integrations.
2019+ consulting (still going)
Been consulting since late 2019, still going strong. Six years in, the work is mostly AI: agent design, RAG pipelines, MCP integrations, and the production hardening that keeps them reliable. Clients have ranged from startups like Gotmy to companies like Santander.
2022-25 Chesskool
2025+ AI engineering
Since mid 2025 I've been building ClassGenie and helping companies ship internal AI systems (RAG pipelines, multi-agent architectures, MCP integrations, and custom evals) that connect their data so their teams can actually use it.